Glucose-sensing by POMC neurons (Parton, Nature 2007) and MCH neurons (Kong, Cell Metabolism, 2010) plays an important role in controlling peripheral glucose homeostasis. Sensing is mediated by a beta-cell like mechanism (glucose - ATP - closure of KATP channels - depolarization), which becomes defective with high fat diet-induced obesity. UCP2 is a likely mediator of obesity-induced loss of glucose sensing. We hypothesize that defective glucose-sensing by these neurons contributes importantly to the pathogenesis of type 2 diabetes. This is being tested in Aim 1 by deleting UCP2 selectively in POMC and MCH neurons. Aims 2-4: Heterogeneity of POMC neurons. POMC neurons regulate both body weight and glucose homeostasis. In general, the field has viewed POMC neurons in the arcuate as a homogenous group - all responding to the same inputs, and all performing the same functions. Alternatively, and likely, there are functionally distinct subsets of POMC neurons, which respond to different inputs and project to different regions of the brain, thus mediating different functions. Of relevance, only a subset of POMC neurons sense glucose, and this capacity is conferred by expression of Suri-containing KATP channels; non-glucose sensing subsets do not express Sur1. Leptin-responding POMC neurons, on the other hand, express leptin receptors (LEPRs), but not Suri. Thus, this proposal hypothesizes that there are two subsets of POMC neurons - glucose-sensing neurons, marked by Suri, which regulate glucose homeostasis, and leptinresponding neurons, marked by LEPRs, which regulate energy balance. In this proposal, genetic approaches and DREADD technology will be used to establish the anatomy (site of projections) and function (glucose- versus body weight-regulating) of these two populations of POMC neurons. These studies should provide new insight into neural mechanisms regulating glucose homeostasis and energy balance.